Hui Guo

Learning and Inference from Complex, Imperfect, and Uncertain Data

Ph.D. Candidate · Department of Computer Science · Western University

Email CV (PDF)

About

I am a fourth-year Ph.D. candidate in the Department of Computer Science at Western University, advised by Prof. Grace Y. Yi and Prof. Boyu Wang.

My research broadly spans statistics and machine learning. I am interested in developing principled methodologies for learning, inference, and decision-making from complex, imperfect, and uncertain data. My work combines ideas from statistical modeling, machine learning, uncertainty quantification, and robust learning.

Broadly, I am interested in probabilistic and latent variable modeling, missing data, causal inference, trustworthy AI, foundation models, and other emerging areas of data-driven discovery. I am particularly interested in applications to healthcare and biomedical research, including multimodal electronic health records, medical imaging, and large language models.

Available for postdoctoral positions starting August 2026.

Research interests

Statistical Machine Learning Learning with Imperfect Data Weak and Crowdsourced Supervision Noisy Label Learning Robust and Distributionally Robust Learning Uncertainty Quantification Medical AI and Foundation Models

Working papers

Hui Guo, Grace Y. Yi (2026).

Instance-Dependent Bayesian Modeling and Robust Training with Crowdsourced Noisy Data. In preparation
Hui Guo, Yi Wan, Boyu Wang, Grace Y. Yi (2026).

Reliability-Aware Longitudinal Clinical Prediction from Multimodal Electronic Health Records. In preparation

Selected publications

Hui Guo, Grace Y. Yi, Boyu Wang (2026).

Addressing both Variable Selection and Misclassified Responses with Parametric and Semiparametric methods.

Bernoulli 32(2): 1303-1327.

paper code arXiv
Gezheng Xu*, Hui Guo*, Li Yi, Charles Ling, Boyu Wang, Grace Y. Yi (2025).

Revisiting Source-Free Domain Adaptation: a New Perspective via Uncertainty Control.

International Conference on Learning Representations (ICLR).

* Equal contribution

paper code poster
Yuan Bian, Yu Shi, Hui Guo, Grace Y. Yi, Wenqing He (2025).

Physician Effects in Critical Care: A Causal Inference Approach Through Propensity Weighting with Parametric and Super Learning Methods.

Journal of Data Science 23(1): 130-148.

paper
Hui Guo, Grace Y. Yi, Boyu Wang (2024).

Learning from Noisy Labels via Conditional Distributionally Robust Optimization.

Neural Information Processing Systems (NeurIPS).

paper code arXiv poster
Hui Guo, Boyu Wang, Grace Y. Yi (2023)

Label Correction of Crowdsourced Noisy Annotations with an Instance-Dependent Noise Transition Model

Neural Information Processing Systems (NeurIPS).

paper code poster

Selected Talks

Learning with Noisy Labels: Variable Selection and Misclassification Probability Modeling.

Lightning talk for Dr. Raymond Carroll's visit · London, Canada · June 2025

slides
Case study: Personalized Physician Recommendation for Critical Care Using the TreeSHAP Method.

Annual Meeting of the Statistical Society of Canada · Online · May 2022

poster